Execution attempt

Sign in with GitHub
← Experiment E2 · Does APT4's tokenization of digits and number separators differ from the Mistral-derived tokenizer's, and does the difference account for the lower GSM8K score of Bielik-PL-11B-v3.0-Instruct and for accuracy differences between English and Polish number formats?

Execution attempt · A8 · planned

Generates the audit number sets S1 to S8 and the GSM8K numeral inventory S9.

Succeeded · Registered by @stw2 via agent. Reported and received times are kept apart; no computation or result is verified.

Pinned source and configuration

https://github.com/stw2/tokenizer-science-tax @ c43f19850803969bc1fefe537caf14e54cfd9a9f

Reference checked 2026-09-14 13:08 UTC. Later commits, branches or plan changes do not retarget this attempt.

Command
python scripts/03_gen_numbersets.py
Working directory
experiments/E02-digit-policy
Configuration paths
None
Parameters
seed 20260703; S2 2,000 log-uniform integers from 10^4 to 10^9; S4 2,000 decimals with 2 to 4 decimal places; six renderings; GSM8K loaded without a revision from a cache holding only 740312add88f781978c0658806c59bc2815b9866; configs/audit_config.json as run had sha256 9af656d3580e51191319ec1c05814d823faf2cf4509df83da11db3f812a388e5, the pinned copy differs only in tokenizers_dir
Environment
CPython 3.12.13 (uv 0.11.16); CPU; Apple silicon, 128 GB unified memory
Output directory
Not recorded

Inputs

Materials the registrant named when registering this attempt, by content identity; obtainability is derived from their location reports. Nothing is fetched or verified.

Delivered events

  1. Registered

    #1

    Generates the audit number sets S1 to S8 and the GSM8K numeral inventory S9.

    reported · received · @stw2 via agent · posted to the Thread

  2. Started

    #2

    Started.

    reported · received · @stw2 via agent · attempt only

  3. Succeeded

    #3

    Completed with exit code 0.

    Exit code 0.

    • s1_integers.jsonl · https://github.com/stw2/tokenizer-science-tax/blob/c43f19850803969bc1fefe537caf14e54cfd9a9f/experiments/E02-digit-policy/data/audit/s1_integers.jsonl · public · sha256 0b7c8db6b7c4… · 588890 bytes · M52
    • s2_biginta.jsonl · https://github.com/stw2/tokenizer-science-tax/blob/c43f19850803969bc1fefe537caf14e54cfd9a9f/experiments/E02-digit-policy/data/audit/s2_biginta.jsonl · public · sha256 ec2979886e7e… · 123932 bytes · M53
    • s3_diagnostics.jsonl · https://github.com/stw2/tokenizer-science-tax/blob/c43f19850803969bc1fefe537caf14e54cfd9a9f/experiments/E02-digit-policy/data/audit/s3_diagnostics.jsonl · public · sha256 2a65e9e8edda… · 2521 bytes · M54
    • s4_decimals.jsonl · https://github.com/stw2/tokenizer-science-tax/blob/c43f19850803969bc1fefe537caf14e54cfd9a9f/experiments/E02-digit-policy/data/audit/s4_decimals.jsonl · public · sha256 9d799e7a5c98… · 122997 bytes · M55
    • s5_renderings.jsonl · https://github.com/stw2/tokenizer-science-tax/blob/c43f19850803969bc1fefe537caf14e54cfd9a9f/experiments/E02-digit-policy/data/audit/s5_renderings.jsonl · public · sha256 e07c5e68d893… · 2915038 bytes · M56
    • s6_units.jsonl · https://github.com/stw2/tokenizer-science-tax/blob/c43f19850803969bc1fefe537caf14e54cfd9a9f/experiments/E02-digit-policy/data/audit/s6_units.jsonl · public · sha256 35c0abb55590… · 55146 bytes · M57
    • s7_signs.jsonl · https://github.com/stw2/tokenizer-science-tax/blob/c43f19850803969bc1fefe537caf14e54cfd9a9f/experiments/E02-digit-policy/data/audit/s7_signs.jsonl · public · sha256 ee93374495b1… · 165225 bytes · M58
    • s8_carriers.jsonl · https://github.com/stw2/tokenizer-science-tax/blob/c43f19850803969bc1fefe537caf14e54cfd9a9f/experiments/E02-digit-policy/data/audit/s8_carriers.jsonl · public · sha256 50049006f91c… · 847472 bytes · M59
    • s9_gsm8k_inventory.json · https://github.com/stw2/tokenizer-science-tax/blob/c43f19850803969bc1fefe537caf14e54cfd9a9f/experiments/E02-digit-policy/data/audit/s9_gsm8k_inventory.json · public · sha256 f655a2e5ad1c… · 6286 bytes · M60
    • data/audit manifest.json · https://github.com/stw2/tokenizer-science-tax/blob/c43f19850803969bc1fefe537caf14e54cfd9a9f/experiments/E02-digit-policy/data/audit/manifest.json · public · sha256 bed188269e5e… · 1363 bytes · M61

    reported · received · @stw2 via agent · posted to the Thread

Report an event

The registrant’s capture tool normally delivers start and outcome events. Reporting here is the same author report with the browser as the reported time; it does not observe the process.

This attempt has a delivered outcome. A new execution is a new attempt.